基于CS和OMP的宽带LFM信号源个数估计算法

摘要

The source number estimation of multi-resource LFM signals is the base of subsequent signal analysis. Considering the compressed measurements which could maintain the feature infor-mation of the original signal, a source number estimation algorithm based on compressed sensing and OMP decomposition is proposed. The algorithm for estimating the source number according the iteration number based on the improved OMP and the iteration termination threshold could be estimated with an adaptive strategy. Experiments results show that the proposed algorithm could obtain high correct rate under the conderation SNR≥-3dB and compression ratio M/N≥0. 25.%多分量LFM信号源个数估计是进行后续信号处理的基础。考虑到压缩测量值较好地保留了原信号的特征信息，基于压缩感知部分重构的思想，提出了一种基于压缩感知和OMP分解的信源个数估计算法，算法在基于OMP的信号重构算法上改进而来，以归一化残差为迭代终止的条件，引入一种自适应阈值选择策略，根据迭代分解次数进行信号源个数估计。仿真结果表明，在信噪比为-3 db左右、压缩比大于0.25的情况下，该算法进行信源个数估计的成功率高于90%。